Tidy Tuesday

Heads up about the way I code:

I like displaying plots in the R Studio console instead of inline (you can switch this setting for each document using the gear)
I’ll try to use double quotations (") instead of single (’), but in case I forget…they’re the same!
Keyboard shortcuts: https://support.rstudio.com/hc/en-us/articles/200711853-Keyboard-Shortcuts
- Run code
  - ctrl/cmd + enter
- %>%
  - ctrl/cmd + shift + m
- Insert code chunk
  - ctrl/cmd + alt + i
- Comment out lines
  - ctrl/cmd + shift + c

# Load packages
library(tidyverse) # readr (read_csv and other data loading functions), dplyr (select, filter, mutate, and other data wrangling functions), magrittr (the piping function), ggplot2 (everything dataviz)
library(plotly) # making interactive graphs (ggplotly())
library(htmlwidgets) # Exporting graphs as HTML (saveWidget())

# Load data
penguins <- read_csv("penguins.csv")

1: IncrediBILL Palmer Penguins

Same data set as last week: Palmer Penguin dataset from LTER data.

Data exploration

# I like to run all these in the console or comment them out after running 

#View(penguins) # Can also click on the object in the global environment
str(penguins) #shows the structure of the dataset

## tibble [344 x 8] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
##  $ species          : chr [1:344] "Adelie" "Adelie" "Adelie" "Adelie" ...
##  $ island           : chr [1:344] "Torgersen" "Torgersen" "Torgersen" "Torgersen" ...
##  $ bill_length_mm   : num [1:344] 39.1 39.5 40.3 NA 36.7 39.3 38.9 39.2 34.1 42 ...
##  $ bill_depth_mm    : num [1:344] 18.7 17.4 18 NA 19.3 20.6 17.8 19.6 18.1 20.2 ...
##  $ flipper_length_mm: num [1:344] 181 186 195 NA 193 190 181 195 193 190 ...
##  $ body_mass_g      : num [1:344] 3750 3800 3250 NA 3450 ...
##  $ sex              : chr [1:344] "male" "female" "female" NA ...
##  $ year             : num [1:344] 2007 2007 2007 2007 2007 ...
##  - attr(*, "spec")=
##   .. cols(
##   ..   species = col_character(),
##   ..   island = col_character(),
##   ..   bill_length_mm = col_double(),
##   ..   bill_depth_mm = col_double(),
##   ..   flipper_length_mm = col_double(),
##   ..   body_mass_g = col_double(),
##   ..   sex = col_character(),
##   ..   year = col_double()
##   .. )

tail(penguins, 5) #shows the last 5 rows of the dataset

## # A tibble: 5 x 8
##   species island bill_length_mm bill_depth_mm flipper_length_~ body_mass_g sex  
##   <chr>   <chr>           <dbl>         <dbl>            <dbl>       <dbl> <chr>
## 1 Chinst~ Dream            55.8          19.8              207        4000 male 
## 2 Chinst~ Dream            43.5          18.1              202        3400 fema~
## 3 Chinst~ Dream            49.6          18.2              193        3775 male 
## 4 Chinst~ Dream            50.8          19                210        4100 male 
## 5 Chinst~ Dream            50.2          18.7              198        3775 fema~
## # ... with 1 more variable: year <dbl>

Static scatterplot

REVIEW: Create a scatterplot using ggplot.

x-axis: body_mass_g
y-axis: bill_length_mm
point color: species

Hint: Use ggplot, geom_point, and aes(color = "")

Tips:

ggplot2 is the package, ggplot is the function
Things ggplot HAS to know: 1) dataset, 2) aesthetics, 3) geom
Dataset and aes can be specified either in ggplot() or any geom_*
Aes is for visual changes that depend on a value in the data (as opposed to visual changes that apply to all values)

ggplot(data = penguins,
       aes(x =  body_mass_g, y = bill_length_mm,
           color = species, shape = sex)) +
  geom_point(na.rm = TRUE) # suppresses warning code about removed NAs

Beautification

Tips:

R Studio cheatsheets: https://rstudio.com/resources/cheatsheets/
Order doesn’t matter… as long as you name your vectors
- When adjusting scales, R will apply your color/shape values in the same order as levels in your data
- If you name the elements of your scale vector after your data values, R will match values instead of using order - this is more foolproof!
Choosing color schemes
- Colors in R: http://www.stat.columbia.edu/~tzheng/files/Rcolor.pdf
- Hex codes: https://coolors.co/ (you can also upload a photo to match colors from)
- Think about data type: continuous, discrete, ordinal (ordered), diverging
Choosing shapes
- In R: http://www.sthda.com/english/wiki/ggplot2-point-shapes
- Unicode: https://jrgraphix.net/r/Unicode/2600-26FF
Adjusting themes
- Changing elements of a theme: https://ggplot2.tidyverse.org/reference/theme.html

# Make the plot pretty!

# Save a vector of colors for use in future graphs
col_scale <- c("#A3E7FC", "#32908F", "#26C485") # c() combines values into a vector
names(col_scale) <- c("Adelie", "Gentoo", "Chinstrap") # naming each value to correspond to the species

# Save a vector of names for use in the legend
lab_scale <- c("P. adeliae", "P. papua", "P. antarcticus") #scientific names :)
names(lab_scale) <- c("Adelie", "Gentoo", "Chinstrap") #ALWAYS name your vectors and double check the order

# New pretty graph
p <- penguins %>% 
  na.omit() %>% # removes ALL lines with NAs from dataset
  ggplot(aes(x = body_mass_g/1000, # can divide by 1000 to make units more manageable
             y = bill_length_mm,
             color = species, # point color depends on species
             shape = sex)) + # point shape depends on sex
  geom_point(size = 1.5, # adjust point size
             stroke = 1.1) + # adjust outline line width
  # scale_color_manual(values = "red", "yellow", "green")) + # DOESN'T WORK
  # scale_color_manual(values = c("lightskyblue", "slateblue4", "darkslategray4")) + 
  scale_color_manual(values = col_scale,
                     labels = lab_scale, # order matters (unless you use a named vector!)
                     name = "Species") +
  # scale_shape_manual(values = c("\u2600", "\u2601")) + # using unicode values
  # scale_shape_manual(values = c("\u2640", "\u2642")) + #another unicode option
  scale_shape_manual(values = c(19, 21), #base R shapes
                     name = "Sex") + #legend title
  theme_minimal() + # one of the default themes
  theme(text = element_text(color = "grey80"), # ALL text color
        title = element_text(face = "bold"), # all TITLES to bold
        plot.title = element_text(color = "lightblue", # PLOT TITLE color
                                  size = 16, # PLOT TITLE text size
                                  hjust = 0.5, vjust = 5), # adjusting title positioning
        plot.margin = margin(25,15,10,10), # remember TRBL (trouble)
        axis.text = element_text(color = "grey70"), # AXIS LABELS color
        legend.text = element_text(face = "italic"), # LEGEND LABELS color
        legend.background = element_rect(fill = "grey25", # LEGEND BACKGROUND color
                                         color = NA), # no legend outline
        plot.background = element_rect(fill = "grey20"), # entire background color
        panel.grid = element_line(color = "grey60")) + # GRID LINE color
  labs(x = "Body mass (kg)", # Make sure units match!!! Especially after converting
       y = "Bill length (mm)",
       title = "IncrediBILL Palmer Penguins")

p #view the named plot

Saving a theme

Like your theme and want to use it for another graph without copy-pasting? Save it!

theme_set replaces the default
theme_update adds new elements to the default theme (analagous to +)
theme_replace replaces elements of the default theme

# What's the default theme in your R session?
ggplot(cars, aes(x = speed, y = dist)) + geom_point()

# Set the default as one of the included themes
theme_set(theme_classic())
#Check it out
ggplot(cars, aes(x = speed, y = dist)) + geom_point()

# Make your own theme!
theme_awful <- theme_dark() # save a copy of a default theme with a new name
theme_set(theme_awful) # set your copy as the default
theme_update(panel.grid.minor = element_line(colour = "red", size = 3), #update your copy!
             plot.background = element_rect(fill = "yellow"),
             title = element_text(size = 60, face = "italic", color = "tan"),
             axis.text = element_text(color = "purple", face = "bold", angle = 160))

#Check your new default theme out
ggplot(cars, aes(x = speed, y = dist)) + geom_point()

Saving the plot

Better to use code than the “export” button
- Can control output size and keep values consistent
- Whatever you do…don’t screenshot! Gives you terrible resolution.

# Check filepath
getwd() # usually run in the console - check where on your computer the plot will be saved

## [1] "C:/Users/carol/Documents/R/tidytuesday-06oct2020"

ggsave("penguin_plot.png", # filepath will be added to WD path (prob just filename) - INCLUDE FILE TYPE EXTENSION.
       plot = p, # name of object to save - default is last_plot()
       height = 4, width = 5, units = "in") # output size (can use inches, cm, pixels)

Interactive scatterplot

Plotly

Plotly is a platform that interacts with R, Python, and Julia
Plotly focuses on creating interactive data visualizations and web applications (check out “Dash”)
We’ll be using ggplotly, which easily turns ggplots into interactive graphs: https://plotly.com/ggplot2/

ggplotly(p) # notice that the plotly graph displays in the "Viewer" pane!

2: Flippin’ Palmer Penguins

Static scatterplot

REVIEW: Create a scatterplot using ggplot.

x-axis: flipper_length_mm
y-axis: body_mass_g
point color: species, same color scheme as above (use the saved vector!)

theme_set(theme_classic()) # switching defaults since I'm not a huge fan of theme_awful...

ggplot(penguins, aes(x = flipper_length_mm, y = body_mass_g,
                     color = species)) +
  geom_point(na.rm=TRUE) + # suppress warning about NAs being removed
  scale_color_manual(values = col_scale) # same color vector as previous graph!

Quick beautification

Tips:

Always make a static graph first - but don’t spend time optimizing all the theme elements!
- some formatting elements do not translate
- legends, element sizes, and themes get funky with ggplotly

The following aesthetic parameters will stay nice when you run your graph through Plotly:

ggplot(penguins, aes(x = flipper_length_mm/10, y = body_mass_g/1000, # adjust units
                           fill = species)) + # changes INSIDE color of the point
  geom_point(na.rm = TRUE,
             shape = 21, # this particular shape allows for a color AND a fill
             color = "grey30") + # changes OUTSIDE color of the point
  scale_fill_manual(values = col_scale, # use scale_fill_* because we used "fill" in the aes
                    name = "Species") +
  labs(x = "Flipper length (cm)", # match units!
       y = "Body mass (kg)", # match units!
       title = "Flippin' Palmer Penguins")

Interactive scatterplot

# Updated static plot
p2 <- ggplot(penguins, aes(x = flipper_length_mm/10, y = body_mass_g/1000,
                           fill = species,
                           cooltip = sex)) + # text to use for tooltip - can have any name
  geom_point(na.rm = TRUE,
             size = 2, shape = 21, color = "grey30") +
  scale_fill_manual(values = col_scale,
                    name = "Species") +
  theme_classic() +
  labs(x = "Flipper length (cm)",
       y = "Body mass (kg)",
       title = "Flippin' Palmer Penguins")

# Interactive plot
ggplotly(p2,
         tooltip = "cooltip") # changes tooltip text. Remember - you can add any data column to the tooltip as long as you name it in aes()

Animated scatterplot

This is ideal for timeseries data - especially measurements that are paired between years.

# Updated static plot
p3 <- ggplot(penguins, aes(x = flipper_length_mm/10, y = body_mass_g/1000,
                           fill = species,
                           cooltip = sex,
                           frame = year)) + # adds animation, with one frame per year
  geom_point(na.rm = TRUE,
             size = 2, shape = 21, color = "grey30") +
  scale_fill_manual(values = col_scale,
                    name = "Species") +
  theme_classic() +
  labs(x = "Flipper length (cm)",
       y = "Body mass (kg)",
       title = "Flippin' Palmer Penguins")

# Interactive plot
p3_move <- ggplotly(p3,
                    tooltip = "cooltip")

p3_move

Saving the plot

Save as HTML

Can save in the HTML output of R markdown
Can save as a standalone HTML file using htmltools

# Save as standalone HTML
saveWidget(p3_move, "penguin_animation.html") #this function is from the htmlwidget package

Export to plotly account

Create a free plotly account: chart-studio.plotly.com
- Find your username and API key
Embed the plot in a webpage: https://help.plotly.com/embed-graphs-in-websites/

# Save plotly username and API key
Sys.setenv("plotly_username" = "katekathrynkat")
Sys.setenv("plotly_api_key" = "WmLiHIWtsIoNpfMz96a6")

# Explort plot to plotly
api_create(x = p3_move, # name of plot
           fileopt = "overwrite", sharing = "public")

See Kate’s embeddable graph: https://chart-studio.plotly.com/~katekathrynkat/39/#/

-Scroll to bottom right -There are 4 icons: Facebook, Twitter, link, <> -The <> icon will bring up the code you can give to a webmaster or embed on an html page

Tidy Tuesday

Pretty Interactive Data Viz with ggplot and plotly

Kate Culhane & Caroline Owens

06 Oct 2020

1: IncrediBILL Palmer Penguins

Data exploration

Static scatterplot

Beautification

Saving a theme

Saving the plot

Interactive scatterplot

2: Flippin’ Palmer Penguins

Static scatterplot

Quick beautification

Interactive scatterplot

Animated scatterplot

Saving the plot